The ACRP annual conference brings together a good cross-section of the clinical research world: regulators, CROs, sponsors, site leaders, and advocates all in the same room for a few days. Our team was there this year, and we came back with a lot to think about.
The honest summary: the environment is shifting faster than many programs are accounting for. Here’s what stood out.
The regulatory environment is stabilizing. Sort of.
If 2025 felt turbulent at FDA (personnel changes, guidance rescissions, some center-level reorganization), 2026 is starting to feel a bit more settled. That’s the good news.
The less straightforward news is that “more stable” doesn’t mean static. FDA’s modernization agenda is actively shaping expectations around AI, real-world data, and adaptive trial design. ICH E6(R3) was formally adopted in September 2025, and sites are still working through what that means in practice. Understanding the changes on paper is one thing. Building processes that actually reflect them is another.
For diagnostic sponsors, this matters because regulatory expectations can shift during the same window you’re designing and executing your study. Programs that don’t account for that risk generating evidence that no longer aligns with what reviewers are looking for by the time you get to submission.
ICH E6(R3) is a bigger deal than some teams realize
Multiple sessions at ACRP addressed the practical implications of GCP’s third revision. The consistent message: R3 isn’t just a terminology update. It’s a meaningful shift in how GCP expects studies to be run.
A few of the changes worth paying attention to:
- Quality is proactive, not reactive. R3 puts the emphasis on preventing errors in critical data and processes, not just documenting deviations after they happen. Quality thinking that brings in key stakeholders earlyneeds to be built into study design from the start, not added as an audit layer later.
- Risk-based monitoring is now the standard. Expect fewer routine site visits and more targeted, issue-driven oversight. That’s a real operational shift in how monitoring resources get allocated and how site problems get escalated.
- Training should be role-based, not a credentialing exercise. Training should correspond to what’s necessary for someone to fulfill their specific role, not blanket requirements that exist just to check a box.
- Less documentation, better documentation. R3 actively discourages “just in case” notation. Source data should be accurate, traceable, and contemporaneous. Not padded.
For diagnostic programs specifically, these aren’t abstract GCP principles. They affect how protocols get built, how sites get qualified, and how the resulting evidence holds up under review.
Study populations are getting more scrutiny
The conversation around enrollment and study populations has been evolving, and it came up again at ACRP this year. The language has shifted somewhat from “diversity” to “representativeness,” but the underlying question remains the same: does your enrolled population reflect the people who will actually use this product?
For IVD and diagnostic device sponsors, that’s not just a philosophical point. It has real implications for site selection, inclusion/exclusion criteria, and how study populations are defined and justified in submissions.
AI in clinical research is moving from conversation to expectation
The AI discussion at ACRP this year felt different from prior years. Less theoretical, more operational. FDA and EMA issued a joint statement aligning on shared AI principles for clinical trials, and FDA released 10 guiding principles for AI in drug development in January 2026, all anchored in human oversight, transparency, and risk-based use.
The takeaway for sponsors isn’t that AI is off-limits. It’s that using AI in a regulated context requires deliberate, documented, and defensible decision-making. That’s worth understanding before you build it into your study operations.
The bigger picture for diagnostic sponsors
The thread running through every session we attended was the same: the clinical research environment is demanding more upstream thinking. Regulatory expectations are showing up earlier in the study lifecycle. Quality has to be designed in, not layered on. And the cost of avoidable design errors keeps going up.
That’s the environment CovarsaDx is built for. We integrate regulatory strategy, clinical execution, and quality oversight from the beginning of a study, not as separate workstreams that come together at submission. When regulatory expectations shift mid-program, our teams can adapt because regulatory thinking is already part of how the study is running.
The sponsors who navigate this well won’t necessarily be the ones who moved the fastest. They’ll be the ones who made the right calls early.
If you’re planning a diagnostic or IVD study and want to talk through how to build it for where the regulatory environment is headed, we’re happy to have that conversation.
Let’s connect: studies@covarsadx.com